Bipeen Acharya '15 Towards an Automated and Customizable Linear Cryptanalysis of a Substitution-Permutation Network Cipher for Use in Embedded Systems

Abstract  The increasing scale of the Internet of Things exposes the proprietors and users to a number of security threats in terms of invasion of privacy, identity thefts and the like.  It is therefore absolutely necessary that the Internet of Things be secure so that communications between these computing devices can happen undisturbed by any illegitimate third party.  Lightweight cryptography has emerged as one of the best solutions to make the communication secure.  Cryptanalysis of a cipher is important to measure and evaluate the strength of a cipher.  In this thesis, I explain in detail the process of linear cryptanalysis on a Substitution Permutation Network of 4 rounds.  I provide an in-depth background of all key concepts and ideas to make this paper accessible to all kinds of readers.  I also provide an open-source Python implementation of the linear cryptanalysis, which can be easily customized and expanded to perform stronger cryptanalysis or to extend on linear cryptanalysis, which can be built upon for further research on cryptanalysis techniques.

Faculty Advisors:  Sara Sprenkle and Serge Salan